APPLICATION NOTE
NUISANCE TRIPPING OF CIRCUIT BREAKERS
(TRUE RMS MEASUREMENTS)
David Chapman
March 2014
ECI Publication No Cu0110
Available from www.leonardo-energy.org
Document Issue Control Sheet
Document Title:
Publication No:
Issue:
Release:
Author(s):
Reviewer(s):
Application Note – Nuisance Tripping of Circuit Breakers (True RMS
Measurements)
Cu0110
03
March 2014
David Chapman
Bruno De Wachter, Stefan Fassbinder
Document History
Issue
Date
Purpose
1
March
2001
Part of the Power Quality Application Guide, published under the name ‘True RMS –
the only true measurement’
2
October
2011
Reworked by the author for adoption in the Good Practice Guide, under the title
‘Nuisance Tripping / (True RMS Measurements)’
3
March
2014
Revision, in the framework of the Good Practice Guide, under the title ‘Nuisance
Tripping of Circuit Breakers / (True RMS Measurements)’
Disclaimer
While this publication has been prepared with care, European Copper Institute and other contributors provide
no warranty with regards to the content and shall not be liable for any direct, incidental or consequential
damages that may result from the use of the information or the data contained.
Copyright© European Copper Institute.
Reproduction is authorised providing the material is unabridged and the source is acknowledged.
Publication No Cu0110
Issue Date: March 2014
Page i
CONTENTS
Summary ........................................................................................................................................................ 1
Nuisance tripping............................................................................................................................................ 2
Inrush Currents ....................................................................................................................................................... 2
Case study ................................................................................................................................................ 2
True RMS – The Only True Measurement .............................................................................................................. 5
What is RMS? ........................................................................................................................................... 6
The consequences of under measurement .............................................................................................. 8
Conclusion ...................................................................................................................................................... 9
Publication No Cu0110
Issue Date: March 2014
Page ii
SUMMARY
Nuisance tripping of circuit breakers is a common problem in many commercial and industrial installations.
This Application Note explains the need to use true RMS measurement instruments when troubleshooting and
analyzing the performance of a power system.
Nuisance tripping of circuit breakers is often caused by the load current being distorted by the presence of
harmonic currents drawn by non-linear loads. Harmonic currents distort the current waveform and increase
the load current required to deliver energy to the load. Many measurement instruments, even quite modern
ones, use an averaging measurement technique that does not measure harmonic currents correctly. The
readings may be as much as 40% too low, and circuit breakers and cable sizes may be underrated as a result.
True RMS meters, which take the complete distorted waveform into account, should be used instead.
Publication No Cu0110
Issue Date:
March 2014
Page 1
NUISANCE TRIPPING
Many commercial and industrial installations suffer from persistent so-called ‘nuisance tripping’ of circuit
breakers. The term refers to the apparently random and inexplicable nature of these events which, although
there is no apparent fault, can cause significant disruption and financial loss. Of course, there is always a
reason for the nuisance tripping of a breaker and there are two common causes. The first possible cause is the
inrush currents that occur when some loads, particularly personal computers and other electronic devices, are
switched on. The second possible cause is that the true RMS current flowing in the circuit has been undermeasured – in other words, the true current really is too high and the trips are valid.
INRUSH CURRENTS
Modern electronic equipment, such as personal computers, monitors, television sets and office equipment,
uses a type of power supply that converts mains electricity to the low voltage direct current without a low
frequency transformer. This type of supply is known as a switched mode power supply (SMPS) and works by
rectifying mains current directly and storing the direct voltage on a large capacitor which charges to the peak
of the supply voltage. Conversion circuits draw current from the capacitor and generate the required low
voltage, usually via a high frequency transformer to provide galvanic isolation. SMPS is very cheap, but it
causes problems in installations because it produces large harmonic currents and draws very large inrush
currents to initially charge the storage capacitor.
Many PCs are never actually ‘turned off’ – in the sense of being electrically isolated from the supply – but
remain powered in standby mode and can be ‘woken up’ by user input, modem activity, or by a network
message. On wake up, they draw a starting current very similar to that drawn when starting from cold.
CASE STUDY
In a small computer room at the University of Sheffield (UK) circuit breakers were being tripped apparently at
random but particularly during the night. Groups of 16 computers were connected to standard final circuits,
each protected by a 32 A miniature circuit breaker. A preliminary investigation revealed nothing – the
installation was apparently correctly installed and functioning correctly. As the problem persisted, further tests
revealed that the trips occurred at the start of the nightly maintenance procedure as the computers turned
back on and it was realized that the inrush current was responsible. Further inspection revealed that the MCBs
were type B devices, so they were replaced with type D devices of the same rating – this will be discussed in
detail later. Although this change resolved the problem, a measurement exercise was undertaken to verify the
conclusion.
A logic controlled switch was interposed at the load side of one of the breakers together with a transient
recorder to measure the applied voltage and current. The switch was capable of applying the supply voltage at
a defined point on the voltage waveform.
A number of startup cycles were conducted and the inrush currents recorded. Figure 1 shows the inrush
current for a typical personal computer and (CRT) monitor set with the supply voltage applied close to the
positive voltage peak. At 155 A, this was the worst case – i.e. the maximum – inrush current recorded during
the testing for this configuration. The voltage waveform is shown only for clarity but it is interesting to note
the resulting distortion of the supply waveform, especially on the first half cycle. The steady state current
consumption was 0.75 A. The current and time resolutions of the measurements are 0.28 A and 0.8 ms
respectively.
Publication No Cu0110
Issue Date:
March 2014
Page 2
Figure 1 – Inrush current for one computer/monitor pair fed by a normal final circuit.
Figure 2 shows the result with 16 computer/monitor pairs connected, the worst case peak current being 583 A.
The current and time resolutions of the measurements in this case are 1.12 A and 0.8 ms respectively.
Figure 2 – Inrush current for 16 computer/monitor pairs fed by a normal final circuit.
Publication No Cu0110
Issue Date:
March 2014
Page 3
MCB CHARACTERISTICS
Although the case study relates to personal computers, the principle applies to most modern electronic
equipment so it is prudent to design installations to survive large inrush currents.
Figure 3 shows the envelope of the characteristic curves for type B, C and D MCBs. The so-called ‘inverse time’
part of the characteristic is designed to protect against over-current. It allows for substantial short-term
overload without tripping, taking advantage of the inherent short time over-current tolerance of the cable. As
the over-current level increases, the time to respond reduces rapidly to restrict the rise in temperature and
reduce the risk of damage. The instantaneous characteristic is intended to respond very rapidly to fault current
to reduce the risk of damage to load circuits.
Types B, C and D MCBs are differentiated by their instantaneous tripping current, shown as B min, Bmax, etc. in
Figure 3.
Figure 3 – Characteristics of type B, C and D MCBs.
Taking the example of a type B MCB, Figure 3 shows that the breaker will not trip ‘instantaneously’ at any
current below 3 times nominal rating, but must trip at or above 5 times nominal. In the case of a nominal 32 A
device, instantaneous tripping will occur between 96 A and 160 A with a type B device, and between 320 A and
640 A for a type D device. The minimum duration required to trip is not well defined.
Publication No Cu0110
Issue Date:
March 2014
Page 4
From the case study results, it is clear that a type B MCB could trip due to the inrush current from one
computer/monitor pair while selecting a type D device would provide a degree of protection against nuisance
tripping. Changing the MCB type means that the potential fault current is increased so the loop impedance of
the final circuit should be carefully checked to ensure compliance with the applicable regulations. In cases
where the inrush is too high for any available breaker, problem loads should be distributed among more final
circuits.
TRUE RMS – THE ONLY TRUE MEASUREMENT
Why do under-measurements occur frequently in modern installations even though digital test instruments
are so accurate and reliable? The answer is that many instruments are not suitable for measuring distorted
currents - and most currents these days are distorted. Since currents are being under-measured, the real
current is much closer to the nominal breaker current than believed, leading to genuine trips that are
misinterpreted as nuisance trips.
This distortion is due to harmonic currents drawn by non-linear loads, especially electronic equipment such as
personal computers, electronically ballasted fluorescent lamps and variable speed drives. Figure 6 shows the
typical current waveform drawn by a personal computer. Obviously, this is not a sine wave and all the usual
sine wave measurement tools and calculation techniques no longer work. This means that, when
troubleshooting or analyzing the performance of a power system, it is essential to use the correct tools for the
job – tools that can deal with non-sinusoidal currents and voltages.
Figure 4 shows two clamp-meters on the same branch circuit. Both the instruments are functioning correctly
and both are calibrated to the manufacturer’s specification. The key difference is in the way the instruments
measure.
Figure 4 - One current, two readings. Which do you trust? The circuit feeds a non-linear load with distorted
current. The True RMS clamp (left) reads correctly but the average responding clamp (right) reads low by 32%.
Publication No Cu0110
Issue Date:
March 2014
Page 5
The left-hand meter is a true RMS instrument and the right-hand one is an averaging reading RMS calibrated
instrument. Appreciating the difference requires an understanding of what RMS really means.
WHAT IS RMS?
The ‘Root Mean Square’ magnitude of an alternating current is the value of equivalent direct current that
would produce the same amount of heat in a fixed resistive load. The amount of heat produced in a resistor by
an alternating current is proportional to the square of the current averaged over a full cycle of the waveform.
In other words, the heat produced is proportional to the mean of the square of the current, so the equivalent
current value is proportional to the root of the mean of the square or RMS. The polarity is irrelevant since the
square is always positive.
For a perfect sine wave, such as that seen in Figure 5, the RMS value is 0.707 times the peak value. Or the peak
value is 2 (= 1.414) times the RMS value. In other words, the peak value of 1 A RMS pure sine wave current
will be 1.414 A. If the magnitude of the waveform is simply averaged (inverting the negative half cycle), the
mean value is 0.636 times the peak, or 0.9 times the RMS value. There are two important ratios shown in
Figure 5:
And
Figure 5 - A pure sine wave.
When measuring a pure sine wave – but only for a pure sine wave – it is quite correct to make a simple
measurement of the mean value (0.636 x peak) and multiply the result by the form factor 1.111 (making 0.707
times peak) and call it the RMS value. This is the approach taken in all analogue meters (where the averaging is
performed by the inertia and damping of the coil movement) and in all older and many currently available
digital multimeters. This technique is described as ‘mean reading, RMS calibrated’ measurement.
The problem is that the technique only works for pure sine waves and pure sine waves do not exist in the real
world of an electrical installation. The waveform in Figure 6 is typical of the current waveform drawn by a
Publication No Cu0110
Issue Date:
March 2014
Page 6
personal computer. The true RMS value is still 1 A, but the peak value is much higher, at 2.6 A, and the average
value is much lower, at 0.55 A.
Figure 6 - Typical waveform of current drawn by a personal computer.
If this waveform is measured with a mean reading, RMS calibrated meter it would read 0.61 A, rather than the
true value of 1 A, nearly 40% too low. Figure 7 below gives some examples of the way the two different types
of meters respond to different wave shapes.
Figure 7 – Response of mean reading and true RMS meters to various waveshapes.
A true RMS meter works by taking the square of the instantaneous value of the input current, averaging over
time and then displaying the square root of this average. Perfectly implemented, this is absolutely accurate
whatever the waveform. Implementation is, of course, never perfect and there are two limiting factors to be
taken into account: frequency response and crest factor.
th
For power systems work it is usually sufficient to measure up to the 50 harmonic, i.e. up to a frequency of
about 2500 Hz. The crest factor, the ratio between the peak value and the RMS value, is important; a higher
crest factor requires a meter with a greater dynamic range and therefore higher precision in the conversion
circuitry. A crest factor capability of at least three is required for accurate measurement in power installations.
It is worth noting that, despite giving different readings when used to measure distorted waveforms, meters of
both types would agree if used to measure a perfect sine wave. This is the condition under which they are
calibrated, so each meter could be certified as calibrated – but only for use on sine waves.
True RMS meters have been available for at least the past 30 years, but they used to be specialized and
expensive instruments. Advances in electronics have now resulted in true RMS measurement capability being
Publication No Cu0110
Issue Date:
March 2014
Page 7
built into many handheld multimeters. Unfortunately, this feature is generally found only towards the top end
of most manufacturers’ ranges, but they are still cheap enough to buy as ordinary instruments for everyone
and every day.
THE CONSEQUENCES OF UNDER MEASUREMENT
The limiting rating for most electrical circuit elements is determined by the amount of heat that can be
dissipated so that the element or component does not overheat.
Cable ratings, for example, are given for particular installation conditions (which determine how fast heat can
escape) and a maximum working temperature. Since harmonic polluted currents have a higher RMS value than
that measured by an averaging meter, cables may have been under-rated and will run hotter than expected;
the result is degradation of the insulation, premature failure and the risk of fire.
Busbars are sized by calculating the rate of heat loss from the bars by convection and radiation and the rate of
heat gain due to resistive losses. The temperature at which these rates are equal is the working temperature
of the busbar, and it is designed so that the working temperature is low enough so that premature ageing of
insulation and support materials does not result. As with cables, errors measuring the true RMS value will lead
to higher running temperatures. Since busbars are usually physically large, skin effect is more apparent than
for smaller conductors, leading to a further increase in temperature.
Other electrical power system components such as fuses and the thermal elements of circuit breakers are
rated in RMS current because their characteristics are related to heat dissipation. This is the root cause of
nuisance tripping – the current is higher than expected so the circuit breaker is operating in an area where
prolonged use will lead to tripping. The response of a breaker in this region is temperature sensitive and may
appear to be unpredictable. As with any supply interruption, the cost of failure due to nuisance tripping can be
high, causing loss of data in computer systems, disruption of process control systems, etc.
Publication No Cu0110
Issue Date:
March 2014
Page 8
CONCLUSION
This paper has described two common causes of nuisance tripping. In each case there are some simple
preventive steps. Avoiding tripping due to inrush currents simply requires selection of the correct type of
breaker and sensible distribution of loads among circuits.
Under-measurement is easily avoided by ensuring that true RMS meters are used routinely. Knowledge of the
real current in a circuit allows corrective action to be taken. For example, redistributing loads across final
circuits.
Publication No Cu0110
Issue Date:
March 2014
Page 9